From a SMILES string to a full developability dossier in seconds — ten fundamental preformulation properties, temperature- and pH-resolved profiles, and interpretable formulation descriptors.
Every module outputs actionable numbers — not just classifications — backed by interpretable ML and confidence estimates.
Ten foundational preformulation properties and a full physicochemical descriptor set from a single SMILES, returned with model confidence and reference ranges.
Density, MP, Tg — bulk & thermallogP, logD₇.₄, logPapp — lipophilicity & permeabilitypKalogS, kinetic solubilityFractionCSP3, TPSA, NumHAcceptors, …
Predict equilibrium solubility under realistic process conditions — pure organics, binary systems, and arbitrary temperatures — without bench screening every combination.
acetone, EtOH, DMSO…)Ionization-aware profiles across the physiological pH range. See how species fraction, intrinsic solubility, and distribution coefficient shift before you commit to a buffer.
logS profilelogD profile
A transparent formulatability score with per-property attribution — every contribution from solubility to permeability is visible, so chemists know exactly what to push on.
An information-rich descriptor set engineered for downstream ML. Combines preformulation predictions with interpretable RDKit features, generated in batch for thousands of compounds.
